The unprecedented urgencies and scope of the HIV epidemic require the rapid development and evaluation of risk behavior change interventions that attempt to promote change in entire vulnerable community populations, not just small groups of individuals who participate in intensive face-to-face interventions. In our current multisite trial, we have established that social diffusion HIV prevention approaches which enlist small cadres of popular opinion leaders to change reference group norms by talking to friends and endorsing behavior change can produce substantial reductions in the risk behavior levels of the larger populations in which the opinion leaders are influential. The current work has been conducted with populations of gay men in 16 small American cities. This renewal application requests support to extend the trial to a new, vulnerable, and understudied community population, women in low-income housing projects. The study will be undertaken in two phases. Phase I (2 1/1 years) will consist of formative studies including elicitation interviews, focus groups, and key informant interviews with housing project women and high-risk heterosexual men. Phase I will also include a diffusion intervention preliminary trial undertaken in 10 low- income housing project populations (five intervention and five comparison). Phase II of the study (2 years) will be a full-scale randomized community field trial testing the model in 20 additional public housing project populations. The intervention tested in the trials will identify, recruit, and engage popular and trusted women in each housing project population to serve as risk reduction diffusion agents to other women in the housing projects; in this model, the opinion leader women will be intensively taught ways to recommend and endorse behavior change strategies to other women and will contract to have risk education conversations with housing project friends and acquaintances. Effects of the intervention on population member behavior will be determined through baseline, postintervention, and follow-up risk interview assessments of all women living in each housing project as well as corroborative indices of population behavior change including condom redemption, STD incidence, HIV testing requests, and reproductive health clinic visits of population members. The overall objective of the work is to identify intervention models that can promote rapid and widescale change in risk behaviors of HIV-vulnerable community populations and to test the intervention in the context of a true randomized community field trial.